Particulars in Context 707As society faces increasingly complexenvironmental and social problems, callsfor integrative, interdisciplinary solu-tions are growing (e.g., Cairns 1993; Patten1994; Lubchenco 1998). As participants in anintegrative discipline, geographers have oftenconsidered themselves ideally situated to tacklethese questions. “Geography has always been(and remains) a generalized—as opposed to aspecialized—discipline. Its viewpoint is one ofbroad understanding,” states one introductoryphysical geography textbook (McKnight 1996:3). Another emphasizes, “Geography is in aunique position to synthesize the environmen-tal, spatial, and human aspects of all these con-cerns” (Christopherson 1997:5). We often makesimilar statements to our students or others—whenever we need to explain our discipline tooutsiders.But privately, among ourselves, we may ques-tion the veracity of these assertions. The bewil-dering multiplicity of subspecialties withingeography makes communication across subdis-ciplinary boundaries difficult, and we wonderwhether, as a discipline, we have carved up theworld to such an extent that we have lost theright to make claims to synthesis or integration.Have we prodigally sold our integrative inheri-tance and run off to squander it on the glitter-ing, reductionist high life?Behind these vague feelings of unease in thediscipline lie valid methodological questions. Inthis paper, I discuss a few of the methodologicalimplications from the vantage point of the sub-discipline of soil geography.1I argue that, despitethe difficulties, we must strive to bridge the gapbetween subdisciplinary reductionist research,on the one hand, and society’s real-world con-cerns on the other. I begin by describing twoapproaches to research in soil geography, thestate-factor approach and the process-basedapproach. The tension between these two orien-tations embodies a key methodological issue insoil geography today. Next, I examine the rela-tive merits of reductionism and synthesis in soilgeography. I conclude by reflecting on thenature of disciplinary identity and the struggle ofintegrative disciplines like soil geography tomaintain a balance between the particulars (i.e.,research involving detailed scales in space andtime) and their overall context (the broad-scaleoutlook generally considered “integrative”).State Factors and Processes:Context and ParticularsAs in the related disciplines of geology andbiology, within soil geography, there exists abasic tension over the relative merits of whatSpedding (1997) terms “(timebound) descrip-tive regional studies” versus “(timeless) analysisof process,” or between what Hoosbeek andBryant (1992) call a “functional” and a “mecha-nistic” understanding. This tension has its rootsin soil geography’s status as a “compositionalhistorical science” focused on changing, individ-ual historical phenomena as well as the proper-ties and processes producing them (Simpson1963; Spedding 1997). Recently, geology andbiology have been converging towards physicsand mathematics, tending to become more law-and process-oriented (Goodwin 1994; Spedding1997).Parallel trends in soil geography are mani-fested in pressures to adopt new methodologicalframeworks for examining the mechanisms andprocesses of soil genesis. In soil geography, time-bound descriptive studies have traditionally uti-lized a functional-factorial or state-factorapproach. In a state-factor model, the soil isviewed in the context of the environmental con-ditions within which soil development occurs.The state factors are not pedogenic processes ormechanisms, but instead are independent vari-ables defining the state of the soil system (Jenny1980; Phillips 1989).The most widespread state-factor model isJenny’s (1941) equation, which relates a soilParticulars in Context: Maintaining a Balance in Soil GeographyLinda R. BarrettDepartment of Geography and Planning, University of Akronproperty (s) to the five independent soil-formingfactors, namely, climate (cl), organisms (o),relief (r), parent material (p), and time (t):s = f (cl, o, r, p, t, ...),but other state-factor models have alsoemerged. For example, Jenny (1961) reformu-lated his original state factors into a new equa-tion with three separate categories of factors:interior conditions, external energy-matterpotentials, and time. Runge’s (1973) energymodel relates soil development to the three fac-tors of organic-matter production, water, andtime. Recognition that factors and processesmay promote either horizonation (progressivepedogenesis) or haploidization (regressive pedo-genesis) in soils led to the development of twoalternative conceptual models, the evolutionmodel of pedogenesis and the dynamic-ratemodel (Johnson and Watson-Stegner 1987;Johnson et al. 1990).State-factor models situate the soil in a par-ticular historical location and thereby provide auseful framework for extrapolating soil charac-teristics across time and space (Amundson andJenny 1997). By carefully selecting study sites,quantifiable relationships between soil proper-ties and the state factors, as determined by thesoil’s location, can be defined using statisticallinear-regression techniques (Richardson andEdmonds 1987). For example, chronofunctionsdescribe change in soil properties with increas-ing surface age (Bockheim 1980; Schaetzl et al.1994). The state-factor model can also be usedqualitatively to assess the stability of a particularcharacteristic or evolutionary trend in the soil(Phillips 1989).Technical shortcomings of the state-factorapproach are numerous. For one, most soils arepolygenetic; i.e., state factors have changedover the course of soil development (Johnson etal. 1990). State factors may also influence soilproperties over a variety of time and spacescales (Phillips 1989). The complexity of theinterrelationships between factors in Jenny’s(1941) formulation of the model suggests that ageneral solution of the equation is not possible(Yaalon 1975). The prospect of complex(chaotic) sensitivity to initial conditions isanother barrier to using statistical regression toquantify a state-factor model (Phillips 1993;Phillips et al. 1996).From a methodological standpoint, however,the state-factor approach appears limited prima-rily because it ignores the mechanisms of soildevelopment. Competing frameworks emphasiz-ing the timeless analysis of process have beenadvanced. For example, Simonson (1959) pro-posed that four